CodeGuide also provide instructions to Cursor Agent to use Supabase MCP to setup backend properly, which is good.
8/ Implementation Plan = Your step-by-step build
This is the most powerful doc in your system.
I write 50+ clear steps to build the app.
Each step is a prompt.
Cursor Agent or Windsurf then builds it like a junior dev, task-by-task.
9/ Run security audits before shipping
To keep your MVP secure by default, I generate a custom security_guideline.mdc doc using @CodeGuidedev
Then inside Cursor or Windsurf:
- Attach this doc to your project
- Switch to Gemini Pro 2.5 (it handles full codebase scans)
- Prompt: “Run a complete security test across the entire codebase. List all vulnerabilities and give clear steps to fix each one.”
It gives you a full audit report before launch, clean, detailed, and dev-ready.
10/ Final Takeaway
AI should never plan your product. That’s your job.
But once you build the right context boundary, it becomes your fastest, cleanest dev copilot.
Build with coding docs. Prompt with clarity. Let AI write the code.
Bookmark this.
It’ll help you for your next project.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
How to actually make your AI-built apps secure (from someone who ships MVPs for a living)
We’ve shipped 45+ AI-built MVPs and learned security the hard way.
If you’re building with Cursor this is how to stay safe 👇
1/ Don’t blindly trust the agent
Cursor flies through code, which is great… until it isn’t.
One of our client projects looked perfect on the surface. @coderabbitai flagged a race condition in the payment system that was actually double-charging users.
The dev thought everything was fine. In production that would have been a disaster.
Let AI write code.
Let another AI review it.
You still approve the changes.
2/ Rate limiting = protecting your wallet
Most AI-built MVPs skip this and then get destroyed by bots.
I know someone whose app got spammed with fake sign-ups and blew through email credits and API usage overnight. The free trial turned into a hefty bill.
Start strict: 100 requests per hour per IP.
Loosen later if needed.
After using @Lovable on 50+ projects, this is how I make sure the UI never looks mid.
AI can design just fine. You are just prompting it wrong.
Here is my actual workflow for shipping clean UI with Lovable.
👇
1/ Start with a reference, not a description
Describing layouts from scratch almost always leads to messy output.
I take a screenshot from Dribbble, drop it into ChatGPT, and ask it to generate a design.json that includes colors, spacing, typography, and layout rules.
Then I tell Lovable to use it for styling only.
This gives me consistent output from the very first screen.
2/ Lock in your colors before you build
Most people open Lovable first and then keep asking it to “fix the UI.”
I spend five minutes on Coolors instead.
I pick a palette, export it, and use those exact values everywhere.
That one decision saves hours of back and forth later.
Cursor + Opus 4.5 is the fastest way to build right now.
But here's the problem no one talks about:
AI-generated code has 1.7x more issues than human-written code.
Here's how to stop shipping broken apps. 👇
1/ The real problem with vibe coding
AI writes clean-looking code. You deploy it. Users find bugs you missed.
1.75x more logic errors. 1.4x more critical issues. 2.74x more XSS vulnerabilities.
The clearer your context, the better your results.
2/ Master the 4 levels of prompting
Level 1: Training Wheels
Use labeled sections in your prompts:
- Context (what you're building)
- Task (what you want)
- Guidelines (how to do it)
- Constraints (what to avoid)
Example:
Bad: "Build me a login page"
Good:
Context: I'm building a SaaS app for small businesses
Task: Create a login page with email/password
Guidelines: Use React, make it mobile-friendly
Constraints: Don't use any external auth services
Structure helps AI understand exactly what you want.
Level 2: No Training Wheels (conversational)
Level 3: Meta Prompting (use AI to improve your prompts)
Level 4: Reverse Meta (document solutions for future use)